Longitudinal　speed　reduction　markings　(LSRMs)　are　designed　to　alert　drivers　to　an　upcoming　change　in　roadway　geometry　(e.g.　direct　connectors　with　smaller　radii).　In　Beijing,　LSRMs　are　usually　installed　on　direct　connectors　of　urban　expressways.　The　objective　of　this　paper　is　to　examine　the　influence　of　LSRMs　on　vehicle　operation　and　driver　behavior,　and　evaluate　the　decelerating　effectiveness　of　LSRMs　on　direct　connectors　with　different　radii.　Empirical　data　were　collected　in　a　driving　simulator,　and　indicators　representing　vehicle　operation　status　and　driving　behavior　were　proposed.　To　examine　the　influence　of　LSRMs,　an　analysis　segment　was　defined,　which　begins　500　m　prior　to　the　entering　point　of　the　connector　and　ends　at　the　exiting　point　of　the　connector.　Furthermore,　the　analysis　segment　was　evenly　divided　into　a　series　of　subsections;　the　length　of　each　subsection　is　50　m.　This　definition　is　introduced　based　on　the　assumption　that　drivers　would　decelerate　smoothly　in　advance　of　the　connector.　The　analysis　results　show　that　drivers　tend　to　decelerate　earlier　when　the　radii　were　200　m　or　300　m.　When　approaching　the　connector,　drivers　tend　to　decelerate　at　500　m　thru　250　m　in　advance　of　the　connector　with　a　200　m　radius;　deceleration　happens　at　300　m-0　m　in　advance　of　the　connector　with　a　300　m　radius.　On　the　connector,　drivers　controlled　the　throttle　pedal　use　at　100　thru　300　m　after　the　entering　point　when　the　radius　was　200　m;　deceleration　occurred　in　two　regions　when　the　radius　was　300　m:　0　m-900　m　from　the　entering　point,　and　the　last　1,000　m　of　the　connector.　The　analytical　results　further　revealed　that　LSRMs　would　be　effective　at　reducing　speeds　when　the　radius　of　the　direct　connector　was　300　m.